Early Complement Activation and Decreased Levels Of

Early Complement Activation and Decreased Levels of Glycosylphosphatidylinositol-Anchored Complement Inhibitors in Human and Experimental Diabetic Retinopathy Jing Zhang, Chiara Gerhardinger, and Mara Lorenzi

Diabetic retinal microangiopathy is characterized by operative in causing retinal vascular damage in diabetes increased permeability, leukostasis, microthrombosis, and how they are triggered are the objects of ongoing and apoptosis of capillary cells, all of which could be investigation. caused or compounded by activation of complement. In Activation of the complement cascade can both com- this study, we observed deposition of C5b-9, the termi- pound and initiate thrombosis, leukostasis, and apoptosis. nal product of complement activation, in the wall of On the one hand, microthrombi and leukostasis can cause ؎ retinal vessels of human eye donors with 9 3 years of ischemia-reperfusion, which activates complement via the type 2 diabetes, but not in the vessels of age-matched nondiabetic donors. C5b-9 often colocalized with von classical pathway (4), and apoptotic endothelial cells can Willebrand factor in luminal endothelium. C1q and C4, trigger the alternative pathway (5). On the other hand, the complement components unique to the classical complement activation is procoagulant, proinflammatory, pathway, were not detected in the diabetic retinas, and proapoptotic (6), and primary activation of comple- suggesting that C5b-9 was generated via the alternative ment on autologous cells can occur because of insufficient pathway, the spontaneous activation of which is regu- activity of plasma and/or cell surface inhibitors. Sponta- lated by complement inhibitors. The diabetic donors neous cleavage of plasma C3 generates C3b, which at- showed a prominent reduction in the retinal levels of taches covalently to the endothelial cell surface through CD55 and CD59, the two complement inhibitors linked to the plasma membrane by glycosylphosphatidylinosi- its reactive thioester group (5). Additional activation steps tol anchors, but not in the levels of transmembrane are normally prevented by complement inhibitors (7). CD46. Similar complement activation in retinal vessels Among these are the cell surface membrane co-factor and selective reduction in the levels of retinal CD55 and protein (MCP; CD46) and decay accelerating factor (DAF; CD59 were observed in rats with a 10-week duration of CD55), which act to restrict the activity of the C3/C5 streptozotocin-induced diabetes. Thus, diabetes causes convertase enzymes, and protectin (CD59), which inhibits defective regulation of complement inhibitors and com- the final step in the assembly of the membrane-attack plement activation that precede most other manifesta- complex (MAC), the terminal and potentially cytotoxic tions of diabetic retinal microangiopathy. These are novel clues for probing how diabetes affects and dam- product of complement activation. Targeted mutation of ages vascular cells. Diabetes 51:3499–3504, 2002 the gene encoding Crry, the rodent complement inhibitor that exhibits both MCP- and DAF-like activity, causes embryonic lethality because of complement activation at the fetomaternal interface (8). Inherited deficiency of he metabolic abnormalities of diabetes cause CD59 in humans causes paroxysmal nocturnal hemoglo- damage to blood vessels throughout the body. binuria (9), and experimental neutralization of CD59 ac- Damage to the retinal capillaries is manifested tivity in rats augments complement-mediated glomerular Twith abnormal permeability that can lead to damage (7). macular edema and with occlusion and obliteration that Glomerular structures (10) and endoneurial micro- can engender retinal ischemia and unregulated angiogen- vessels (11) of patients with diabetes show signs of esis. Underlying processes documented to date are micro- complement activation. Decreased availability or effec- thrombosis (1), leukostasis (2), and accelerated apoptosis tiveness of complement inhibitors in diabetes is suggested of vascular cells (3). Whether these are the only processes by the findings that high glucose in vitro selectively decreases on the endothelial cell surface the expression of CD55 and CD59 (12), the two inhibitors that are glyco- From the Schepens Eye Research Institute and Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts. sylphosphatidylinositol (GPI)-anchored membrane pro- Address correspondence and reprint requests to Mara Lorenzi, Schepens teins (7), and that CD59 undergoes nonenzymatic Eye Research Institute, 20 Staniford St., Boston, MA 02114. E-mail: [email protected]. glycation that hinders its complement-inhibitory function Received for publication 26 June 2002 and accepted in revised form 16 (13). In this work, we investigated whether complement August 2002. activation is a feature of human nonproliferative diabetic DAF, decay accelerating factor; GPI, glycosylphosphatidylinositol; mAb, monoclonal antibodies; MAC, membrane attack complex; MCP, membrane retinopathy and is associated with changes in inhibitory co-factor protein; RBC, red blood cell; vWf, von Willebrand factor. molecules. We extended the study to an animal model of

DIABETES, VOL. 51, DECEMBER 2002 3499 COMPLEMENT IN THE DIABETIC RETINA

TABLE 1 Characteristics of eye donors and specimens Eyes: time to Eyes: time to Age Sex Diabetes duration enucleation processing Groups (years) (M/F) (years) (h) (h)* Immunohistochemical studies Nondiabetic 66 Ϯ 6 8/6 4 Ϯ 113Ϯ 4 Diabetic 67 Ϯ 6 6/7 9 Ϯ 33Ϯ 112Ϯ 5 Protein studies Nondiabetic 64 Ϯ 7 14/1 3 Ϯ 131Ϯ 6 Diabetic 66 Ϯ 7 11/6 9 Ϯ 34Ϯ 332Ϯ 7 Data are means Ϯ 1 SD. *The time elapsed from death to retina processing was longer for the eyes used for protein isolation (fresh samples) than for the eyes used in the morphological studies because the latter were fixed by the eye banks before shipment. nonproliferative diabetic retinopathy to test the universal- mAb MEM-43 (1 ␮g/ml) and goat polyclonal Ab N-20 (1:1,000; Santa Cruz). ity and timing of abnormalities. Neuron-specific enolase was detected with mAb BBS/NC/VI-H14 (1 ␮g/ml; Dako). In the rat retinas, CD46, CD55, and CD59 were detected with rabbit polyclonal antibody H-294 (1:1,000), H319 (1:200), and R-79 (1:1,000), respec- RESEARCH DESIGN AND METHODS tively (all from Santa Cruz), and ␤-actin with mouse mAb AC-15 (1:200,000; Eye donors and specimens. Human postmortem eyes were obtained from Sigma, St. Louis, MO). certified eye banks through the National Disease Research Interchange. To determine whether the antibodies chosen could detect CD59 quantita- Criteria for inclusion in the study were age Ͻ75 years, diabetes duration Ͻ15 tively in diabetic samples irrespective of glycation, we tested the antibodies years to address mostly nonproliferative retinopathy, and the fewest possible against red blood cells (RBCs) CD59 glycated in vitro. Normal human RBCs chronic pathologies other than diabetes. Criteria for exclusion were retinal, were incubated in standard storage medium containing 55 mmol/l glucose or hematological, and inflammatory diseases; uremia; chemotherapy; and use of in storage medium containing 150 mmol/l glucose for 40 h at 30°C, following life support measures. The causes of death were almost exclusively cardio- a protocol previously described (17). Glycation of hemoglobin was measured vascular in both groups, and hypertension was reported in one-third of donors with the Glyc-Affin GHb Assay (Perkin Elmer Wallac, Norton, OH). Equal in each group. The eyes used for immunohistochemical studies were from 13 amounts (200 ␮g) of solubilized protein from the high- and control-glucose diabetic and 14 nondiabetic donors (Table 1). The eyes were fixed in 10% RBC were incubated with mouse mAb BRA-10G anti-human CD59 (20 ␮g/ml; buffered formalin by the eye banks; one retina from each donor was Ancell, Bayport, MN) at 4°C overnight. The immunocomplexes were precipi- cryopreserved and embedded in OCT compound for sectioning, and the other tated by the addition of protein-G Sepharose (Sigma) for 2 h, and the resulting was in most cases digested with trypsin to isolate the intact microvascular immunoprecipitates were analyzed by Western blotting (14). CD59 glycation, network (trypsin digests) (3). The retinal trypsin digests of seven donors in known to lead mostly to formation of Amadori product (13), was measured each group had previously been studied for microvascular cell apoptosis and with a polyclonal antibody (provided by N. Taniguchi) that reacts with histological lesions of nonproliferative diabetic retinopathy (3). The eyes used ⑀-(1-deoxyhexitolyl)-lysine, the reduced form of the Amadori product (18). for protein isolation were from 17 diabetic and 15 nondiabetic donors Hence, the blots were treated with NaHBH4 (50 ␮mol/l in PBS) for4hatroom (Table 1). temperature, followed by a 15-min wash in PBS acidified to pH 5.0 with acetic Animals and specimens. Six-week-old male Sprague-Dawley rats (Taconic acid to stop the reaction. After additional PBS washes, the membrane was Farms, Germantown, NY) were assigned randomly to a diabetic and a control blocked and reacted with the anti–hexitol-lysine antibody (1:1,000). The group. Diabetes was induced by intravenous administration of streptozotocin amount of immunoprecipitated CD59 was measured in companion blots (55 mg/kg body wt, dissolved in citrate buffer pH 4.5). The care of the diabetic probed with the anti-CD59 mAb MEM-43 or polyclonal antibody N-20. Similar rats and insulin treatment to prevent weight loss were as described previously immunoprecipitation experiments were performed with lysates of individual (14). The rats were killed after 10 weeks of diabetes. A total of nine diabetic retinas of diabetic and nondiabetic donors to assess the level of glycation of and nine control rats were studied; from each rat, one retina was embedded retinal CD59 in human diabetes. in OCT to prepare frozen sections, and the other was homogenized in lysis Statistical analysis. Data are summarized as the mean Ϯ SD. Statistical buffer for isolation of protein. analysis was performed with the unpaired t test. Immunohistochemistry. Sections (10 ␮m) and trypsin digests from the formalin-fixed human retinas were rehydrated and blocked as described (15) RESULTS and incubated overnight at 4°C with the primary antibodies diluted in PBS containing 2% BSA and 0.3% Triton X-100. The reactions were visualized by MAC deposition in human diabetic retinal vessels. peroxidase immunohistochemistry or immunofluorescence (15,16). The pri- Radial retinal sections from 12 of the 13 diabetic donors mary antibodies were mouse monoclonal antibodies (mAb) aE11 anti-human studied showed C5b-9 immunostaining. The staining local- C5b-9 (0.5 ␮g/ml; Dako, Carpinteria, CA), which react with a neoepitope on ized at small and mid-size vessels (Fig. 1), which were poly C9 exposed upon formation of MAC; 1A4 anti-human smooth muscle actin (1.8 ␮/ml; Dako); E4.3 anti-human CD46 (1 ␮g/ml; Santa Cruz Biotech- mostly negative for smooth muscle actin. In contrast, no nology, Santa Cruz, CA); MEM-43 anti-human CD59 (1 ␮g/ml; Serotec, Oxford, staining or only occasional punctiform staining was ob- U.K.); and rabbit polyclonal antibodies to human C3d (1:1,000), C1q (1:1,000), served in vessels of nondiabetic donors. Also, the retinal C4c (1:1,000), and von Willebrand factor (vWf; 1:7,000), all from Dako. The rat capillaries, not readily visualized in sections and therefore retinal sections were fixed in acetone and studied with mAb 2A1 anti-rat C5b-9 (1:150; provided by W.G. Couser). Negative controls were obtained by studied in trypsin digest preparations, showed MAC immu- substituting the primary antibodies with an equivalent concentration of nostaining in the diabetic but not nondiabetic donors (Fig. nonimmune isotypic mouse IgG or rabbit IgG as appropriate. 1). In diabetic retinal vessels observed in confocal micros- Immunoblotting and immunoprecipitation. For isolation of protein, the copy, MAC was seen on the endothelial surface, often fresh retinas were homogenized in ice-cold lysis buffer containing phospha- colocalized with vWf (Fig. 2). Occasionally, MAC was also tase and protease inhibitors as described (16). The homogenate was sonicated three times for 2 s and centrifuged at 16,000g for 15 min at 4°C, and the detected in the middle and outer layers of the wall of large supernatant was collected and stored in aliquots at Ϫ80°C. Protein concen- vessels, concentric to the lumen and possibly reflecting tration was determined with the Bradford method using BSA as standard cumulation over time in basement membranes (Fig. 2). (protein assay kit; Bio-Rad, Hercules, CA). Retinal proteins were resolved by MAC colocalizes with C3 but not with complement SDS-PAGE and immunoblotted as described (14). For the study of CD46, CD55, and CD59, SDS-PAGE was under nonreducing conditions. In the human components of the classical pathway. In the retinal retinas, CD46 was detected with rabbit polyclonal antibody H-294 (1:1,000; sections of diabetic donors, MAC showed perfect colocal- Santa Cruz), CD55 with mAb BRIC 216 (1 ␮g/ml; Serotec), and CD59 with both ization with C3, the complement component on which the

3500 DIABETES, VOL. 51, DECEMBER 2002 J. ZHANG, C. GERHARDINGER, AND M. LORENZI

FIG. 1. MAC in retinal vessels of diabetic donors. A–D: In retinal sections immunostained for C5b-9, the brown product of the peroxidase reaction is present in the vessels of two diabetic donors (A and C; arrows) but not in vessels of nondiabetic donors (B and D). E and F: Retinal trypsin .␮m 40 ؍ digests show C5b-9 immunofluorescence in the capillaries of a diabetic (E) but not in those of a nondiabetic (F) donor. Bar classical and alternative pathways of activation converge. mass; its levels were reduced by 48% in the diabetic retinas In confocal microscopy, colocalization was captured on when compared with the control retinas (P ϭ 0.001). CD59 the surface of vascular endothelium (data not shown). In migrated as a broad band at 20–25 kDa, reflecting a the same retinas, C1q and C4, the two complement com- mixture of glycoforms (19), and its levels measured using ponents unique to the classical pathway, were either not either Ab MEM-43 or N-20 were decreased as well in the detected at all or detected occasionally in vascular lumens retinas of diabetic donors (P ϭ 0.01). CD46 showed in the reflecting their presence in blood (Fig. 3). This pattern is retina the two isoforms (65 and 55 kDa) reported in other consistent with complement activation occurring in dia- organs and accounted for by alternative splicing, leading betic retinal vessels via the alternative pathway. to different degrees of glycosylation (20). An additional Selective decrease in GPI-anchored complement in- band migrating at 48 kDa possibly represents intracellular hibitors in diabetic retinal vessels. Because activation precursors (20). In contrast to CD55 and CD59, which are of the alternative pathway of complement is critically anchored to the plasma membrane by GPI linkage, CD46 is modulated by inhibitors, we compared the levels of inhibitors in the retina of diabetic and nondiabetic donors and investigated their topography in the human retina. In SDS-PAGE and Western blot (Fig. 4A), CD55 was detected as a unique band of 70- to 80-kDa apparent molecular

FIG. 2. MAC localization within retinal vessels of diabetic donors. In FIG. 3. Absent C1q and C4 staining in the retina of diabetic donors. retinal sections immunostained for both C5b-9 (red fluorescence) and Consecutive retinal sections from a diabetic donor were immuno- vWf (green fluorescence), C5b-9 (arrows) is detected on the vascular stained for C5b-9 (A, red fluorescence) and C1q (B, green fluores- endothelium (A) and in deeper layers of the wall of a large vessel (B) cence); sections from a different donor were stained for C5b-9 (C) and .C4 (D). Vessels are positive for C5b-9 (arrows) but not for C1q or C4 ؍ and in several areas colocalizes with vWf (yellow, arrowheads). Bars .␮m 40 ؍ ␮m. Bars 5

DIABETES, VOL. 51, DECEMBER 2002 3501 COMPLEMENT IN THE DIABETIC RETINA

this protein is a target of nonenzymatic glycation in diabetes. Exposure of human RBCs to 150 mmol/l glucose in vitro resulted in nonenzymatic glycation of hemoglobin (7.4 vs. 5.1% in control RBCs incubated in 55 mmol/l glucose) and of CD59 (Fig. 4B). However, the intensity of the CD59 band as detected by mAb MEM-43 or polyclonal antibody N-20 was similar in the immunoprecipitate from RBCs exposed to 55 or 150 mmol/l glucose, indicating that documented glycation of CD59 did not affect its reactivity with the two antibodies (Fig. 4B). Confirmation was obtained in immunoblots comparing the levels of CD59 in glycated and nonglycated RBC lysates not subjected to immunoprecipitation (data not shown). There was a sug- gestion that CD59 was more glycated in diabetic than in nondiabetic retinas (the ratio of the hexitol-lysine to the CD59 signal was 1.3 Ϯ 0.8 in the diabetic and 0.7 Ϯ 0.4 in the nondiabetic retinas), but the difference was not statis- tically significant. We did not find antibodies suitable for immunohistochemical detection of CD55 in the fixed human retina, but both CD59 and CD46 could be localized exclusively to vessels (Fig. 5). Retinal vessels thus seem to be the site where the selective decrease in GPI-anchored complement inhibitors occurs in diabetes. MAC deposition and selective decrease in GPI- anchored complement inhibitors in the retinal vessels of diabetic rats. Radial retinal sections from rats with 10 weeks of diabetes (GHb levels 16 Ϯ 4% vs. 4 Ϯ 1% in controls) showed vascular deposition of MAC, similar to that observed in the retina of human diabetic donors (Fig. 6A). No MAC immunostaining was found in the retina of nondiabetic rats. Also in agreement with the findings in human diabetic donors were the decreased levels of CD59 and CD55 but not CD46 in the retina of diabetic rats when compared with levels in nondiabetic rats (Fig. 6B).

FIG. 4. Complement inhibitors in the retina of diabetic donors. A, Left: DISCUSSION Whole retinal lysates (20 ␮g per lane) from diabetic (D) and nondiabetic (ND) donors were resolved by SDS-PAGE under nonreducing We found that retinal vessels, one of the main targets of conditions, and the blots were probed with antibody to complement the long-term effects of diabetes, show evidence of com- inhibitors (CD55, CD59, CD46). Sequential probing with antibody to plement activation early in diabetes, associated with a neuron-speciﬁc enolase (NSE) was used as control for loading (16). Right: Quantitation of the densitometric signals; the bars represent prominent and selective decrease in the levels of GPI- -mean ؎ SD of the values obtained in the indicated number of patients. anchored complement inhibitors. This indicates that com *P < 0.01 versus ND. B: Lysates of human RBCs exposed in vitro to control (C) or high (H) glucose concentrations were immunoprecipi- plement regulation is altered in diabetes and suggests tated with antibody to CD59 and resolved by SDS-PAGE. The blots mechanisms for the pathogenesis of diabetic microangi- were probed with antibody to hexitol-lysine to test CD59 glycation and opathy. with the two CD59 antibodies used to compare CD59 levels in diabetic and nondiabetic retinas. The presence of MAC in diabetic retinal vessels is likely to reﬂect local assembly, possibly on the endothelium. a transmembrane protein, and its levels were similar in Confocal images captured colocalization of MAC and vWf, diabetic and nondiabetic retinas (the sum of the three which, coupled with the knowledge that the C5b-9 com- bands was used to measure CD46 levels in each sample). plex becomes partially embedded within the plasma mem- The measurement of CD59 levels took into account that brane of target cells (21), indicates that vascular

FIG. 5. Localization of complement inhibitors in the human retina. In retinal sections from a nondiabetic donor, the brown product of the peroxidase reaction shows CD59 (A) and CD46 (B) immunostaining only in the vessels. C: Negative control. The arrow points to the same large .␮m 40 ؍ vessel, probably a vein, in the consecutive sections. Bars

3502 DIABETES, VOL. 51, DECEMBER 2002 J. ZHANG, C. GERHARDINGER, AND M. LORENZI

FIG. 6. MAC and complement inhibitors in the retina of rats with 10-week duration of streptozotocin-diabetes. A: Retinal sections from a control and a diabetic rat show C5b-9 immunostaining (green fluorescence) localized to .␮m 50 ؍ vessels only in the diabetic rat (arrows). Bars B: Complement inhibitors were studied by immunoblot in whole retinal lysates from control (C) and diabetic (D) rats as described in the legend to Fig. 4 for the human retinal lysates. The loading control was ␤-actin; its levels were similar in the control and diabetic rat retinas. Bars represent mean ؎ SD of the densitometric readings recorded in the indicated number of rats. *P < 0.01 versus controls. endothelium is one of the sites of MAC assembly in number of apoptotic cells detected at any given time in the diabetic retinal vessels. However, in larger retinal vessels retinal vessels of diabetic donors (3) and preceded by of diabetic donors we also observed MAC deposits re- several months the development of acellular capillaries in moved from the luminal endothelium. Increased perme- the diabetic rats (3). The decreased levels of inhibitors ability of retinal vessels in diabetes cannot be the only also cannot be readily attributed to complement activation explanation for MAC deposited in the vessel wall because because the latter results in upregulation of inhibitors (7). there was no evidence of extravasation of complement The reduced levels of CD55 and CD59 thus may be viewed components (we tested C1q and C4), which have molecu- as a primary effect of diabetes and one of the mechanisms lar sizes smaller than MAC. Moreover, when complement for complement activation in diabetic vessels. is activated by the subendothelial extracellular matrix, the The selective decrease in GPI-anchored complement reaction seems not to proceed to MAC formation (22). In inhibitors suggests effects of diabetes on common regula- the renal glomeruli of patients with diabetes, immunoelec- tory steps in the synthesis or the processing of these tron microscopy had shown that MAC is deposited on molecules. Studies in cultured endothelial cells have membranous structures, most likely cellular debris shown that, similar to what we observed in diabetes, high trapped between layers of existing and newly formed glucose concentrations decreased cellular content of both lamellae of basement membranes (10). Reduplication of CD55 and CD59 but not CD46 and that the levels of basement membranes is a characteristic feature of dia- GPI-anchored inhibitors decreased on the cell surface but betic vessels (23). It is thus possible that MAC forms increased in the culture medium (12). Transfer to a initially on the luminal endothelium of diabetic vessels and glucose medium induced in yeast spheroplasts the activa- that, over time, fragments of disrupted plasma membrane tion of an endogenous GPI-specific phospholipase C (26). still bearing MAC become embedded within adjacent Collectively, these observations propose that a mechanism basement membranes. Such a paradigm would be consis- for the selectively reduced levels of CD55 and CD59 in tent with the absence of MAC deposits in the wall of the diabetic retinal vessels may be sought in hyperglycemia- larger retinal vessels of diabetic rats, which were studied induced activity of phospholipases capable of cleaving GPI after only 10 weeks of diabetes. anchors. Hyperglycemia can additionally compromise the The observations in the rat model identify complement protective effects of CD59 (and other complement inhibi- activation and decreased levels of inhibitors as early tors) by causing nonenzymatic glycation in the vicinity of events in the course of diabetes, detectable before other the active site (13). In this context, one may ask why manifestations of retinal vascular pathology. The human patients who have diabetes with poor metabolic control do retinal vessels were from donors with a duration of not manifest symptoms of paroxysmal nocturnal hemoglo- diabetes that had set the stage also for the occurrence of binuria. Complete deficiency of CD59 is probably required vascular cell apoptosis (3) and microthrombosis (1), both for the syndrome to become apparent (9). In addition, the of which could contribute to complement activation (4,5). mechanism that leads to decreased levels of CD59 (and However, these two processes are not detected in the CD55) in diabetes may be cell- or tissue-specific and not diabetic rat model until after 6–8 months of diabetes involve RBCs. It was noteworthy in our experiments that (3,14,24), and this points to complement activation as an the RBCs exposed to high glucose in vitro manifested upstream event. Decreased levels of complement inhibi- extensive Amadori adduct formation on CD59 but did not tors seem likewise to be upstream of other known vascu- show decreased CD59 levels. lar pathology. Complement regulators are lost from This work proposes several unsuspected pathways necrotic cells and tissues (25), but the prominent decrease whereby diabetes/hyperglycemia may alter the phenotype in CD55 and CD59 was disproportionate to the small of vascular cells to induce the characteristic abnormalities

DIABETES, VOL. 51, DECEMBER 2002 3503 COMPLEMENT IN THE DIABETIC RETINA of retinal microangiopathy (27). C3a and C5a anaphylatox- cause of paroxysmal nocturnal hemoglobinuria. N Engl J Med 323:1184– ins generated during complement activation could contrib- 1189, 1990 10. Falk RJ, Sisson SP, Dalmasso AP, Kim Y, Michael AF, Vernier RL: ute to increased permeability and neutrophil adhesion. Ultrastructural localization of the membrane attack complex of comple- Even if not lytic, repeated attacks by MAC will, at a ment in human renal tissues. Am J Kidney Dis 9:121–128, 1987 minimum, burden the endothelium with the energetic and 11. Rosoklija GB, Dwork AJ, Younger DS, Karlikaya G, Latov N, Hays AP: metabolic cost involved in recovering from the attacks Local activation of the complement system in endoneurial microvessels of diabetic neuropathy. Acta Neuropathol 99:55–62, 2000 (28) and may lead to a state of chronic activation (6), 12. Accardo-Palumbo A, Triolo G, Colonna-Romano G, Potestio M, Carbone M, heightening the risk of acute events such as apoptosis and Ferrante A, Giardina E, Caimi G, Triolo G: Glucose-induced loss of microthrombosis over time. The finding that diabetes glycosyl-phosphatidylinositol-anchored membrane regulators of comple- alters the synthesis or processing of GPI-anchored pro- ment activation (CD59, CD55) by in vitro cultured human umbilical vein teins further widens the spectrum of implications for the endothelial cells. Diabetologia 43:1039–1047, 2000 13. Acosta J, Hettinga J, Flu¨ ckiger R, Krumrei N, Goldfine A, Angarita L, endothelial phenotype to include possible effects on other Halperin J: Molecular basis for a link between complement and the GPI-linked molecules and on signaling systems compart- vascular complications of diabetes. Proc Natl Acad SciUSA97:5450– mentalized in the plasmalemmal microdomains in which 5455, 2000 GPI-anchored molecules cluster (26,29). New experiments 14. Gerhardinger C, McClure KD, Romeo G, Podesta` F, Lorenzi M: IGF-I mRNA and signaling in the diabetic retina. Diabetes 50:175–183, 2001 will address these issues to clarify further the biology of 15. Gerhardinger C, Brown LF, Roy S, Mizutani M, Zucker CL, Lorenzi M: diabetic vascular disease and to test whether the events Expression of vascular endothelial growth factor in the human retina and surrounding complement activation and/or complement nonproliferative diabetic retinopathy. Am J Pathol 152:1453–1462, 1998 activation itself should become targets in preventative 16. Podesta` F, Romeo G, Liu W-H, Krajewski S, Reed JC, Gerhardinger C, strategies. Lorenzi M: Bax is increased in the retina of diabetic subjects and is associated with pericyte apoptosis in vivo and in vitro. Am J Pathol 156:1025–1032, 2000 17. Cagliero E, Roth T, Roy S, Lorenzi M: Characteristics and mechanisms of ACKNOWLEDGMENTS high-glucose-induced overexpression of basement membrane components This work was supported by National Institutes of Health in cultured human endothelial cells. Diabetes 40:102–110, 1991 Grant EY 09122, the Juvenile Diabetes Research Founda- 18. Myint T, Hoshi S, Ookawara T, Miyazawa N, Suzuki K, Taniguchi N: Immunological detection of glycated proteins in normal and streptozoto- tion Center for Diabetic Retinopathy at the Schepens Eye cin-induced diabetic rats using anti hexitol-lysine IgG. Biochim Biophys Research Institute, and the George and Frances Levin Acta 1272:73–79, 1995 Endowment. 19. Rudd PM, Morgan BP, Wormald MR, Harvey DJ, van den Berg CW, Davis We thank Dr. William Couser for mAb 2A1 and Dr. SJ, Ferguson MAJ, Dwek RA: The glycosylation of the complement Naoyuki Taniguchi for the antibody to hexitol-lysine. regulatory protein, human erythrocyte CD59. J Biol Chem 272:7229–7244, 1997 20. Post TW, Liszewski MK, Adams EM, Tedja I, Miller EA, Atkinson JP: Membrane cofactor protein of the complement system: alternative splicing REFERENCES of serine/threonine/proline-rich exons and cytoplasmic tails produces 1. Boeri D, Maiello M, Lorenzi M: Increased prevalence of microthromboses multiple isoforms that correlate with protein phenotype. J Exp Med in retinal capillaries of diabetic individuals. Diabetes 50:1432–1439, 2001 174:93–102, 1991 2. Schro¨ der S, Palinski W, Schmid-Scho¨ nbein GW: Activated monocytes and 21. Ware CF, Kolb WP: Assembly of the functional membrane attack complex granulocytes, capillary nonperfusion, and neovascularization in diabetic of human complement: formation of disulfide-linked C9 dimers. Proc Natl retinopathy. Am J Pathol 139:81–100, 1991 Acad SciUSA78:6426–6430, 1981 3. Mizutani M, Kern TS, Lorenzi M: Accelerated death of retinal microvascu- 22. Hindmarsh EJ, Marks RM: Complement activation occurs on subendothe- lar cells in human and experimental diabetic retinopathy. J Clin Invest lial extracellular matrix in vitro and is initiated by retraction or removal of 97:2883–2890, 1996 overlying endothelial cells. J Immunol 160:6128–6136, 1998 4. Walport MJ: Complement at the interface between innate and adaptive 23. Vracko R: Basal lamina layering in diabetes mellitus: evidence for accel- immunity. N Engl J Med 344:1140–1144, 2001 erated rate of cell death and cell regeneration. Diabetes 23:94–104, 1974 5. Tsuji S, Kaji K, Nagasawa S: Activation of the alternative pathway of 24. Ishibashi T, Tanaka K, Taniguchi Y: Platelet aggregation and coagulation in human complement by apoptotic human umbilical vein endothelial cells. the pathogenesis of diabetic retinopathy in rats. Diabetes 30:601–606, 1981 J Biochem 116:794–800, 1994 25. Va¨keva¨ A, Laurila P, Meri S: Loss of expression of protectin (CD59) is 6. Tedesco F, Fischetti F, Pausa M, Dobrina A, Sim RB, Daha MR: Comple- associated with complement membrane attack complex deposition in ment-endothelial cell interactions: pathophysiological implications. Mol myocardial infarction. Lab Invest 67:608–616, 1992 Immunol 36:261–268, 1999 26. Mu¨ ller G, Grob E, Wied S, Bandlow W: Glucose-induced sequential 7. Nangaku M: Complement regulatory proteins in glomerular diseases. processing of a glycosyl-phosphatidylinositol-anchored ectoprotein in Sac- Kidney Int 54:1419–1428, 1998 charomyces cerevisiae. Mol Cell Biol 16:442–456, 1996 8. Xu C, Mao D, Holers VM, Palanca B, Cheng AM, Molina H: A critical role 27. Lorenzi M, Gerhardinger C: Early cellular and molecular changes induced for murine complement regulator Crry in fetomaternal tolerance. Science by diabetes in the retina. Diabetologia 44:791–804, 2001 287:498–501, 2000 28. Morgan BP: Effects of the membrane attack complex of complement on 9. Yamashina M, Ueda E, Kinoshita T, Takami T, Ojima A, Ono H, Tanaka H, nucleated cells. Curr Top Microbiol Immunol 178:115–140, 1992 Kondo N, Orii T, Okada N, Okada H, Inoue K, Kitani T: Inherited complete 29. Anderson RGW: The caveolae membrane system. Annu Rev Biochem deficiency of 20-kilodalton homologous restriction factor (CD59) as a 67:199–225, 1998

3504 DIABETES, VOL. 51, DECEMBER 2002